The Engine

This chapter documents the structure itself — not that a person can be made to comply, but the specific, short, transferable procedure that breaks a mind down until compliance is the only move left. It is not temper, or cruelty, or lost control. It was written down, as a numbered list, by the United States government, in 1957.

In that year the sociologist Albert Biderman published a study of how Chinese and North Korean interrogators had extracted false confessions from captured American airmen during the Korean War. What Biderman found was not exotic torture but a compact, reproducible method — the Chart of Coercion, a list of eight techniques: isolation; monopolization of perception; induced debility and exhaustion; threats; occasional indulgences; demonstrating omnipotence and omniscience; degradation; and the enforcement of trivial demands.¹ The psychological syndrome those techniques produce had been named that same year by a separate research team — the psychologists I. E. Farber, Harry Harlow, and Louis Jolyon West, who reduced it to three words: debility, dependency, and dread. Sixteen years later, Amnesty International reproduced Biderman’s chart in its own Report on Torture, describing these eight techniques as the universal tools of torture and coercion — universal because they recur, unchanged, wherever the deliberate breaking of a human being is practiced, regardless of who is doing it or why.²

Read the eight in sequence and their logic becomes visible: not eight separate cruelties but eight coordinated moves toward one objective — destroying the target’s ability to check reality against any source the controller does not command. Isolation removes every other person who could confirm what is true. Monopolization of perception fills the vacuum, making the controller the only source of information about the world and about the target themselves. Induced exhaustion — through lost sleep, drained resources, relentless low-grade crisis — strips away the mental reserve that resistance requires. Threats install a permanent background of dread. Occasional indulgences, the unpredictable kindness after the cruelty, are the cruelest instrument of all, because they manufacture hope and gratitude and bind the target to the very person harming them. The demonstration of omnipotence and omniscience, I always know where you are; there is no point resisting, makes defiance feel not merely dangerous but pointless. Degradation raises the internal cost of holding on to a self worth defending. And the enforcement of petty, shifting, pointless demands drills in the habit of obeying until obedience no longer needs to be commanded. The end state is a person who can no longer independently verify what is happening to them, and no longer believes resistance is possible. Every other tactic lands more easily on a mind prepared this way; this is why the engine comes first.

Biderman’s chart matters to a book mostly not about prisoners of war because of the equivalence it exposes. The same eight techniques that describe a Korean interrogation cell describe, with almost no translation, an ordinary controlling marriage — which is why the Chart of Coercion is now a standard teaching tool in domestic-violence advocacy, handed to shelter workers and police to recognize the pattern with no war and no cell anywhere in the frame. The mapping is to coercively controlling relationships specifically, not to ordinary marital conflict — a distinction domestic-violence researchers themselves insist on, and one this chapter returns to below.

Isolation from friends and family; control of what a partner is allowed to read, hear, and believe; the exhaustion of never having a settled moment; the threat that no one would believe her anyway; the good week that arrives just as she is about to leave; I always know where you are; the humiliation; the endless petty rules about the house and the schedule and the tone of voice. Nothing has to be added to Biderman’s list to describe a home, and nothing removed. At the smallest scale, before any state or syndicate enters the account, the demonstration is already complete: the mechanism that breaks a mind does not change when the operator changes. Only its resources do.

If the grammar is structural — eight coordinated moves toward one end — then it cannot depend on the sex of the person running it, any more than a lever depends on the hand that pulls it. This is not a hedge appended for balance; it is a straightforward consequence of the thesis. The point is easy to state and easy to forget, because almost every account, including much of this one, defaults to a controlling man and a woman in his grip. Nothing in the machine’s construction requires that arrangement. In 2018 Jordan Worth became the first woman convicted under the coercive-control offence created by the Serious Crime Act 2015; she had isolated her male partner, Alex Skeel, from family and friends, seized control of his finances and communications, starved him, and beaten him nearly to death, and she was jailed for seven and a half years.³ The same offence reads identically across a same-sex relationship. A documented female perpetrator and a male victim do not weaken the invariance claim. They are a demonstration of it — of exactly one part of it, as what follows will insist.

That settles one question and leaves a larger one standing, and the difference between them matters enough to be spelled out rather than glossed. It is one thing to say the machine does not require a man to run it; Jordan Worth demonstrates that, and it is not seriously in dispute. It is a different thing to say that the gendered arrangement of ordinary life has little to do with why the machine works so well inside a house — and Evan Stark, whose apparatus this book borrows throughout, argues precisely the opposite. On his account, coercive control gets its purchase from a persisting sexual inequality that hands the operator his levers pre-built: the economic dependency, the settled assumption about whose work is whose, the credibility gap already waiting in a courtroom. A single documented female perpetrator does not touch that argument, any more than a single female chief executive refutes a gendered labor market. The two claims are not the same claim, and only the first of them is answered by Jordan Worth.

So the position here is narrower than a denial, and it is the one the rest of this book actually demonstrates. The machine does not need gender. It needs an inequality — some asymmetry already standing, already load-bearing, already legitimate enough that leaning on it draws no attention. Inside a marriage that asymmetry is usually the gendered one, and on that scale Stark is describing something real. Change the scale and the machine simply reaches for whatever else is lying there. For the Thai farmworkers of Global Horizons it was immigration status: recruitment debt engineered to be unrepayable, passports confiscated, and a visa that tied them to one employer, so that leaving the job meant leaving the country.⁴ In a hospital it is clinical authority. At the state scale, as Part VII will show, it is clearance and the classification stamp. The engine is indifferent to which inequality it metabolizes; it requires only that one be available. That is a stronger claim than the one it replaces, not a weaker one, because it says what the machine is looking for rather than merely what it can do without — and it is why the same parts assemble, unchanged, in a marriage and in a black site.

What the engine produces is the condition the previous chapter’s psychology already named — learned helplessness, the state in which a person who could act stops trying, having been taught by experience that action changes nothing.⁵ This corrects the most common misunderstanding about coercion. The engine’s purpose is not primarily to inflict pain. Pain is only one of its instruments, and often not the most effective one. Its purpose is to produce dependency and helplessness, because a dependent, helpless person can be influenced with a fraction of the force an intact one would require. Breaking the mind is not the alternative to persuasion. It is the preparation for it.

A fair objection has been building since the start of this chapter: if isolation, monitoring, exhaustion, and petty rules describe coercive control, do they not also describe a strict parent, a demanding employer, a hard training regime, or simply an intense marriage? If the engine describes everything, it explains nothing. The distinction is real, and it is positive rather than a matter of degree. The sociologist Michael Johnson drew it most usefully by separating what he called intimate terrorism — violence and behavior embedded in an overarching pattern of coercive control and domination — from situational couple violence, the conflict-driven friction of a relationship in which neither party is trying to govern the other.⁶ The mark of coercive control is never the individual act, which will always resemble ordinary strictness or an ordinary bad day. It is the presence of an overarching structure whose object is the target’s subjugation — the systematic stripping of another person’s autonomy and liberty. Legitimate authority is bounded, accountable, and oriented toward the subordinate’s development or a shared task: the parent aims at a child who will one day not need parenting, the trainer at a soldier who can act alone. Coercive control is unbounded, unaccountable, and oriented toward a person who will need permission indefinitely. The individual acts can be identical. Only the purpose tells them apart.

Honesty requires pressing on that distinction, because serious people doubt it can carry the weight the law has begun to place on it. A body of legal scholarship — most comprehensively the comparative survey edited by Marilyn McMahon and Paul McGorrery — argues that coercive control, once written into criminal statute, is dangerously hard to bound: its definitions differ from one jurisdiction to the next, it names a pattern rather than a discrete act, and a pattern is far harder to prove to a criminal standard than a bruise.⁷ The record bears the worry out. Since England and Wales created the offence of controlling or coercive behaviour in 2015, recorded cases have climbed year on year while convictions have lagged well behind — the same diffuseness that makes the abuse feel total to the person inside it also makes it resist the kind of proof a court demands.⁸ And there is a sharper form of the objection still: an offence broad enough to describe a whole relationship is broad enough to be turned around, wielded by the more plausible party against the one it was written to protect — the reversal this book documents in family courts later on.

None of that is answered by insisting the phenomenon is real, and this book will not pretend it is. Three replies are owed, plainly. The standard urged here is not the courtroom’s but reconstruction from external, corroborable traces — a higher and narrower bar than the intuitive impression the critics rightly fear, and one built to fail safely when the traces are not there. The worry that the framework can be turned into a weapon is not the critics’ alone; it is this book’s own, argued at length in the chapters on the manufactured case, and the reply to it is that same discipline rather than a denial that the danger is real. And what finally divides intimate terrorism from an unhappy marriage is not only the controlling structure, which an argument can always dispute, but the specific, independently diagnosable injury it leaves behind — a clinical signature no merely rhetorical category could produce on cue, in unrelated case after unrelated case. The construct is contestable. The wound is not.

---

The eight techniques are fixed. The channels through which they are delivered are not, and they have multiplied enormously — which is why a procedure first documented in a 1957 study of prison camps needs restating now. Three modern delivery vehicles are worth documenting directly. Each runs Biderman’s engine through a medium he never anticipated, and each is attested by court records, government admissions, and the operators’ own conduct.

The first is sound. Music delivered at sufficient volume and duration is not a metaphor for coercion; it is monopolization of perception and induced debility administered acoustically, and it has been used as such, openly, by the United States. In December 1989, when General Manuel Noriega took refuge in the Vatican’s diplomatic mission in Panama City, American forces surrounded the building and bombarded it around the clock with deafening rock music — a documented psychological operation that ended only after the Holy See formally complained, three days later.⁹ What was played as almost a joke against a dictator was, within a few years, standardized as an interrogation technique. The musicologist Suzanne Cusick’s study of music in the “war on terror” documented its deliberate use to destroy a detainee’s sense of an interior, private self — to leave, as she put it, nowhere to hide.¹⁰ The clearest single case is Mohammed al-Qahtani, Guantánamo detainee 063, whose leaked interrogation log records loud music used to prevent sleep across twenty-hour interrogations, alongside prolonged isolation, forced nudity, and sexual humiliation. Here the record is a concession, not an allegation: Susan Crawford, the senior official overseeing the military commissions, stated on the record that what was done to al-Qahtani met the legal definition of torture.¹¹ Sound leaves no wound to photograph, which is precisely its administrative appeal — it is the engine run through a medium that produces no evidence of itself.

The second is the feed. A networked platform can deliver threats, degradation, and the demonstration of omnipotence at a scale and relentlessness no single person could sustain, while dissolving responsibility across so many hands that no individual appears to have done anything actionable. The most thoroughly adjudicated case is Alex Jones’s decade-long campaign against the families of the children killed at Sandy Hook Elementary School, whom he repeatedly told his audience were actors in a staged hoax. His followers acted on it, subjecting bereaved parents to years of harassment, threats, and confrontation, in person and online, until some were driven into hiding. The juries settled the characterization: Connecticut and Texas juries in 2022 found Jones liable for defamation and intentional infliction of emotional distress, returning verdicts totalling roughly one and a half billion dollars, and the underlying findings survived every appeal.¹² The mechanism is Biderman’s exactly — surround the target, make the hostility feel total and inescapable, degrade them past the point where a self feels defensible — but the medium is a broadcast feed, and the deniability is structural: the man who supplied the narrative never sent a single threat himself, and each follower who did was only one voice among thousands. The same architecture operates at interpersonal scale through coordinated online harassment and stalking, documented later in this book in cases that ended in criminal conviction rather than civil verdict.¹³

The third is the deniable threat delivered in person. Here the engine is run not louder but quieter — engineered so the target cannot prove anything happened at all, which turns Biderman’s fourth technique, threats, into something closer to a private hallucination the victim cannot corroborate. The documented template is the East German Stasi’s practice of Zersetzung, literally “decomposition,” a program of psychological destruction that deliberately avoided arrest in favor of tactics designed to leave no evidence: entering a target’s home to move or replace small objects, spreading rumors through their own social circle, orchestrating a stream of minor misfortunes calibrated to make the target doubt their own perception and to make any complaint sound paranoid. It is examined in full at state scale later in this book; what matters here is that it is not folklore but an admitted, file-documented government program.¹⁴ The same grammar appears wherever organized coercion needs to threaten without a prosecutable record — the warning phrased as concern, the accident that could have happened to anyone, the message whose menace is entirely in a context only the target and the sender share. The deniability is not a side effect. It is the design, and it is the through-line that connects a whispered threat in a kitchen to a state security service’s most refined work.

Across all three channels the invariant holds. Whether the instrument is a fist, a song played on a loop, a broadcast that turns strangers into a mob, or a rearranged shelf meant to make someone doubt their own memory, the eight techniques underneath are the same eight. The rest of this book is, in one sense, a catalog of delivery vehicles — the devices, institutions, budgets, and tradecraft that carry the engine to larger and larger targets. The engine itself does not change from the first page to the last. It only gets better resourced, and better hidden. Each of the three cases here also shows the limit of that hiding: the conceded torture, the verdicts that survived every appeal, the deniable program now read out of the Stasi’s own surviving files. Deniability by design bought each operator years, not permanence.

Notes

1. Albert D. Biderman, “Communist Attempts to Elicit False Confessions from Air Force Prisoners of War,” Bulletin of the New York Academy of Medicine 33, no. 9 (1957): 616–625, in which Biderman set out the eight interrogation techniques published as the “Chart of Coercion.” The three-word syndrome those techniques produce — debility, dependency, and dread (the “DDD” triad) — was named in the same year by I. E. Farber, Harry F. Harlow, and Louis Jolyon West, “Brainwashing, Conditioning, and DDD (Debility, Dependency, and Dread),” Sociometry 20, no. 4 (December 1957): 271–285.↑

2. Amnesty International, Report on Torture (London: Duckworth / Amnesty International Publications, 1973), which reproduced Biderman’s chart and characterized the eight techniques as the universal tools of torture and coercion.↑

3. In April 2018, at Luton Crown Court, Jordan Worth pleaded guilty to controlling or coercive behaviour in an intimate relationship, wounding with intent, and causing grievous bodily harm, and was sentenced to seven and a half years’ imprisonment for the prolonged, near-fatal abuse of her partner, Alex Skeel — isolation from family and friends, control of his finances and communications, starvation, and violence. She was the first woman convicted under section 76 of the Serious Crime Act 2015, the offence of controlling or coercive behaviour in an intimate or family relationship (in force December 2015, England and Wales), which applies without regard to the sex or sexual orientation of either party; the offence therefore reads identically in same-sex relationships. In June 2018 the Court of Appeal declined to increase the sentence under the unduly-lenient-sentence scheme. Widely reported; see BBC News coverage of the case and the Bedfordshire Police account.↑

4. The Thai-farmworker facts are established in EEOC v. Global Horizons, Inc. (D. Haw.), the U.S. Equal Employment Opportunity Commission’s largest farmworker human-trafficking case: recruitment fees engineered into unrepayable debt, passports confiscated on arrival, and visas tying workers to a single employer. In December 2014 the court entered findings of fact and conclusions of law against Global Horizons and awarded damages to the Thai claimants; the company has since ceased operations. U.S. EEOC, EEOC v. Global Horizons (2014).↑

5. The learned-helplessness research (Martin Seligman and Steven Maier, 1967) is treated in Chapter One; it is named again here because it identifies the specific psychological end-state Biderman’s eight techniques are designed to produce, and explains why a broken target can be influenced with far less force than an intact one.↑

6. Michael P. Johnson, A Typology of Domestic Violence: Intimate Terrorism, Violent Resistance, and Situational Couple Violence (Boston: Northeastern University Press, 2008). Johnson distinguishes intimate terrorism — violence and coercion embedded in a general pattern of control and domination — from situational couple violence, which is conflict-driven and is not embedded in an overarching effort to control the partner. The presence or absence of that controlling structure, rather than the severity of any single incident, is what separates the two, and it is that criterion — not the individual act — that distinguishes coercive control from bounded, accountable authority.↑

7. Marilyn McMahon and Paul McGorrery, eds., Criminalising Coercive Control: Family Violence and the Criminal Law (Springer, 2020) — a cross-jurisdictional examination that catalogues the definitional variability, evidentiary difficulty, and risk of misuse against victims raised by critics of the offence.↑

8. The offence of controlling or coercive behaviour in an intimate or family relationship was created by section 76 of the Serious Crime Act 2015 (England and Wales, in force December 2015). Recorded offences rose year on year thereafter while the share resulting in conviction fell after the first year — a gap attributed largely to the difficulty of evidencing a sustained course of conduct to the criminal standard. UK Home Office, Review of the Controlling or Coercive Behaviour Offence (2021); Office for National Statistics, coercive-control offence data (2016–2025).↑

9. Operation Nifty Package, December 1989. When Manuel Noriega took refuge in the Apostolic Nunciature (the Holy See’s diplomatic mission) in Panama City following the U.S. invasion, American forces subjected the building to continuous loud rock music as psychological pressure; the Holy See complained to the U.S. government, and the music was discontinued after roughly three days. Widely documented in contemporaneous reporting and subsequent military histories.↑

10. Suzanne G. Cusick, “Music as Torture / Music as Weapon,” Trans: Revista Transcultural de Música 10 (2006). Cusick documents the deliberate use of high-volume music in detention settings in the “global war on terror” and argues that its function is to destroy the prisoner’s sense of a private, interior self.↑

11. The interrogation log of Mohammed al-Qahtani (Guantánamo detainee 063), leaked and published in 2006, records loud music used to prevent sleep during interrogations lasting up to twenty hours, alongside 160 days of isolation, forced nudity, and sexual and religious humiliation. Susan J. Crawford, the convening authority for the Guantánamo military commissions, stated on the record that al-Qahtani’s treatment met the legal definition of torture — the first such admission by a senior Bush-administration official. Bob Woodward, “Detainee Tortured, Says U.S. Official,” The Washington Post, January 14, 2009.↑

12. In 2022, juries in Connecticut (Lafferty v. Jones, October 2022) and Texas (the Heslin and Lewis trial, August 2022) found Alex Jones liable for defamation and intentional infliction of emotional distress arising from his repeated claims that the 2012 Sandy Hook Elementary School shooting was a staged hoax; the Connecticut award, with punitive damages, reached approximately $1.4 billion and the Texas award roughly $49 million. Trial testimony established years of harassment and threats against the victims’ families by Jones’s audience. The U.S. Supreme Court declined to disturb the Connecticut judgment in 2025.↑

13. See the treatment of coordinated online harassment and cyberstalking in Part II, including cases resulting in federal criminal conviction, where the same surround-and-degrade architecture operates against a single private target.↑

14. Zersetzung (“decomposition”) — the East German Ministry for State Security’s program of deniable psychological destruction — is documented from the Stasi’s own surviving files and examined in full at state scale in Chapters 38 and 43. It is named here as the documented template for the deniable, evidence-free threat, not as an isolated example.↑